The Study:

Bernard Erlin, of Erlin, Himes Associates, performed petrographic analysis on six core samples of failed swimming pool plaster from Florida in 1986-1987.

• Core 2 was from a plaster finish showing streaks and blotches.
• Core 3 was a plaster finish described as containing dark discolorations flanking fine cracks, with pocked circular areas up to ¾ inch in diameter.
• Core 5 was taken from a plaster showing affected areas ¼ - 3/8 inch in diameter, which were also found under the tile line.
• Core 6 was taken from a plaster finish with dime size spots of discoloration, that occurs also along cracks and were prominent throughout the pool.

Their Conclusions:

Core 2: "Aggregate particles are exposed on the finished surface due to a loss of a thin overlying "paster" layer. Rimming some of these particles is soft paste. In one localized surface area, paste over aggregate particles is stained pale yellow. Paste does not evidence carbonation. In the deteriorated areas are relict (relic) portland cement particles and marble particles having the fineness of the cement.

Core 3: "Surface is smooth, except where pocked by circular areas up to ¾ inch in size and up to 3/32 inch deep. These areas are discolored brownish grey.
Paste of these areas is very soft and has a chalky texture. Relict cement particles are present, along with marble particles have the fineness of the cement. Unusual is the absence of carbonated portland cement paste, a normally inevitable phenomenon for paste on concrete surfaces. In the bottom of these localized deteriorated areas, paste is carbonated.
The deterioration begins as a small area associated with an aggregate particle and progresses laterally and vertically.

Core 5 and 6 "The surface is marred by many deteriorated and pale yellow discolored areas up to 3/8 inch in size. Paste of the deteriorated areas is very soft, and not carbonated. The deteriorated paste has a chalky texture. Marble particles having the same fineness of the paste, plus relict portland cement particles, are present in the deteriorated areas.

Erlin suggests a precarbonation of the paste may be a method for restricting the deterioration, using either negative Langelier index water, or creating a high CO2 atmosphere (5 to 7 percent) in a tented pool. (This recommendation was incorporated into Whitney's work.) Furthermore, the final paragraph of Erlin's summary and discussion section of his report states: "The exact nature of the chemical mechanism responsible for the surface deterioration is not known, a rare phenomenon in this area." (The insinuation is that aggressive chemical attacks are well understood by petrographers and material scientists, but the conditions evident on these plaster cores are not the result of this obvious kind of chemical action.)